Full-stack developer with a focus on AI-powered products and multi-agent systems. I design and build agent-based architectures, ship user-facing apps and APIs, and bring product sense to engineering, from prototype to production. I work well across product and engineering and prefer practical solutions that reduce complexity for users and teams.
My GitHub account showcases most of my public work. Unfortunately, many of my contributions are to private repositories.
https://github.com/Astronaut828
Open-source profile (projects + tutorials)
https://speedrunethereum.com/builders/0x992AB820a2776Ff79284B9844D338B5C5fc792e9
Technical documentation translation (German/English) — security-focused developer docs.
https://ethereum.org/de/developers/docs/smart-contracts/security/
📧 Email: [email protected]
💻 Github: Astronaut828
🔗 LinkedIn: Marco Carrino
🌎 Remote (US) — Austin, TX
🇩🇪 German
🇺🇸 English
Aug.26: I am building and operating several live products that demonstrate practical AI-assisted engineering in production. Brave Daily, an automated SMS product for military families, reached 57 subscribers in its first two weeks and now runs through a production pipeline covering payments, onboarding, scheduled content delivery, retries, monitoring, and operational recovery. ToddlerTips extends this work into a fully automated visual-content pipeline, while DadJokes serves as an automated content-generation and publishing system. Together, these projects show how AI can support real products with active users, not only by generating content or code, but by participating in the operational systems that keep those products running.
My broader focus is developing agent-based workflows across the full software development lifecycle. Using Codex, Claude, and local agent systems, I coordinate AI across product discovery, architecture, implementation, debugging, testing, review, content operations, and production monitoring. The work is centered less on one-off prompting and more on building repeatable, verifiable human-in-the-loop systems that can carry meaningful engineering work from an early idea into a deployed and actively used and paid product. I am especially interested in the judgment layer: managing context, designing effective agent boundaries, verifying outputs, responding to production failures, and understanding where automation still requires human editorial or technical oversight.
Feb.26: I am building a local-first AI systems lab around OpenClaw, local models, and personal automation workflows. The goal is to move beyond collecting ideas and into a working environment where I can run agents, coordinate tasks, inspect runtime state, and tighten the loop between planning, execution, memory, and review. This work includes experimenting with multi-agent orchestration, local model workflows, session monitoring, and Mission Control-style dashboards that show what is actually running on the machine in real time. Long term, I want this setup to become a practical foundation for turning notes, prototypes, and half-formed product ideas into usable tools, while keeping the system private, fast to iterate on, and grounded in real execution instead of abstract planning.
Stack: Next.js, TypeScript, Postgres, Redis, Stripe, OpenAI API
Dad Jokes Today - Fresh Dad Jokes Every Day
Stack: Next.js, TypeScript, Postgres, Redis, Vercel Cron, OpenAI API